Barrie Slaymaker

NAME

XML::Filter::Dispatcher - Path based event dispatching with DOM support

SYNOPSIS

    use XML::Filter::Dispatcher qw( :all );

    my $f = XML::Filter::Dispatcher->new(
        Rules => [
            'foo'               => \&handle_foo_start_tag,
            '@bar'              => \&handle_bar_attr,

            ## Send any <foo> elts and their contents to $handler
            'snarf//self::node()'  => $handler,

            ## Print the text of all <description> elements
            'description' 
                    => [ 'string()' => sub { push @out, xvalue } ],
        ],

        Vars => {
            "id" => [ string => "12a" ],
        },
    );

DESCRIPTION

WARNING: Beta code alert.

A SAX2 filter that dispatches SAX events based on "EventPath" patterns as the SAX events arrive. The SAX events are not buffered or converted to an in-memory document representation like a DOM tree. This provides for low lag operation because the actions associated with each pattern are executed as soon as possible, usually in an element's start_element() event method.

This differs from traditional XML pattern matching tools like XPath and XSLT (which is XPath-based) which require the entire document to be built in memory (as a "DOM tree") before queries can be executed. In SAX terms, this means that they have to build a DOM tree from SAX events and delay pattern matching until the end_document() event method is called.

Rules

A rule is composed of a pattern and an action. Each XML::Filter::Dispatcher instance has a list of rules. As SAX events are received, the rules are evaluated and one rule's action is executed. If more than one rule matches an event, the rule with the highest score wins; by default a rule's score is its position in the rule list, so the last matching rule the list will be acted on.

A simple rule list looks like:

    Rules => [
        'a' => \&handle_a,
        'b' => \&handle_b,
    ],

Actions

There are several types of actions:

  • CODE reference

        Rules => [
            'a' => \&foo,
            'b' => sub { print "got a <b>!\n" },
        ],
  • SAX handler

        Handler => $h,  ## A downstream handler
        Rules => [
            'a' => "Handler",
            'b' => $h2,    ## Another handler
        ],
  • undef

        Rules => [
            '//node()' => $h,
            'b' => undef,
        ],

    Useful for preventing other actions for some events

  • Perl code

        Rules => [
            'b' => \q{print "got a <b>!\n"},
        ],

    Lower overhead than a CODE reference.

    EXPERIMENTAL.

EventPath Patterns

Note: this section describes EventPath and discusses differences between EventPath and XPath. If you are not familiar with XPath you may want to skim those bits; they're provided for the benefit of people coming from an XPath background but hopefully don't hinder others. A working knowledge of SAX is necessary for the advanced bits.

EventPath patterns may match the document, elements, attributes, text nodes, comments, processing instructions, and (not yet implemented) namespace nodes. Patterns like this are referred to as "location paths" and resemble Unix file paths or URIs in appearance and functionality.

Location paths describe a location (or set of locations) in the document much the same way a filespec describes a location in a filesystem. The path /a/b/c could refer to a directory named c on a filesystem or a set of e<ltc>> elements in an XML document. In either case, the path indicates that c must be a child of b, b must be <a>'s, and <a> is a root level entity. More examples later.

EventPath patterns may also extract strings, numbers and boolean values from a document. These are called "expression patterns" and are only said to match when the values they extract are "true" according to XPath semantics (XPath truth-ness differs from Perl truth-ness, see EventPath Truth below). Expression patterns look like string( /a/b/c ) or number( part-number ), and if the result is true, the action will be executed and the result can be retrieved using the xvalue method.

TODO: rename xvalue to be ep_result or something.

We cover patterns in more detail below, starting with some examples.

If you'd like to get some experience with pattern matching in an interactive XPath web site, there's a really good XPath/XSLT based tutorial and lab at http://www.zvon.org/xxl/XPathTutorial/General/examples.html.

Actions

Two kinds of actions are supported: Perl subroutine calls and dispatching events to other SAX processors. When a pattern matches, the associated action

Examples

This is perhaps best introduced by some examples. Here's a routine that runs a rather knuckleheaded document through a dispatcher:

    use XML::SAX::Machines qw( Pipeline );

    sub run { Pipeline( shift )->parse_string( <<XML_END ) }
      <stooges>
        <stooge name="Moe" hairstyle="bowl cut">
          <attitude>Bully</attitude>
        </stooge>
        <stooge name="Shemp" hairstyle="mop">
          <attitude>Klutz</attitude>
          <stooge name="Larry" hairstyle="bushy">
            <attitude>Middleman</attitude>
          </stooge>
        </stooge>
        <stooge name="Curly" hairstyle="bald">
          <attitude>Fool</attitude>
          <stooge name="Shemp" repeat="yes">
            <stooge name="Joe" hairstyle="bald">
              <stooge name="Curly Joe" hairstyle="bald" />
            </stooge>
          </stooge>
        </stooge>
      </stooges>
    XML_END
Counting Stooges

Let's count the number of stooge characters in that document. To do that, we'd like a rule that fires on almost all <stooge> elements:

    my $count;

    run(
        XML::Filter::Dispatcher->new(
            Rules => [
                'stooge' => sub { ++$count },
            ],
        )
    );

    print "$count\n";  ## 7

Hmmm, that's one too many: it's picking up on Shemp twice since the document shows that Shemp had two periods of stoogedom. The second node has a convenient repeat="yes" attribute we can use to ignore the duplicate.

We can ignore the duplicate element by adding a "predicate" expression to the pattern to accept only those elements with no repeat attribute. Changing that rule to

                'stooge[not(@repeat)]' => ...

or even the more pedantic

                'stooge[not(@repeat) or not(@repeat = "yes")]' => ...

yields the expected answer (6).

Hairstyles and Attitudes

Now let's try to figure out the hairstyles the stooges wore. To extract just the names of hairstyles, we could do something like:

    my %styles;

    run(
        XML::Filter::Dispatcher->new(
            Rules => [
                'stooge' => [
                    'string( @hairstyle )' => sub { $styles{xvalue()} = 1 },
                ],
            ],
        )
    );

    print join( ", ", sort keys %styles ), "\n";

which prints "bald, bowl cut, bushy, mop". That rule extracts the text of each hairstyle attribute and the xvalue() returns it.

The text contents of elements like <attitudes> can also be sussed out by using a rule like:

                'string( attitude )' => sub { $styles{xvalue()} = 1 },

which prints "Bully, Fool, Klutz, Middleman".

Finally, we might want to correlate hairstyles and attitudes by using a rule like:

    my %styles;

    run(
        XML::Filter::Dispatcher->new(
            Rules => [
                'stooge' => [
                    'concat(@hairstyle,"=>",attitude)' => sub {
                        $styles{$1} = $2 if xvalue() =~ /(.+)=>(.+)/;
                    },
                ],
            ],
        )
    );

    print map "$_ => $styles{$_}\n", sort keys %styles;

which prints:

    bald => Fool
    bowl cut => Bully
    bushy => Middleman
    mop => Klutz

Examples that need to be written

  • Examples for accumulating data

  • Advanced pattern matching examples

Sending Trees and Events to Other SAX Handlers

When a blessed object $handler is provided as an action for a rule:

    my $foo = XML::Handler::Foo->new();
    my $d = XML::Filter::Dispatcher->new(
        Rules => [
           'foo' => $handler,
        ],
        Handler => $h,
    );

the selected SAX events are sent to $handler.

Element Forwarding

If the event is selected is a start_document() or start_element() event and it is selected without using the start-document:: or start-element:: axes, then the handler ($foo) replaces the existing handler of the dispatcher ($h) until after the corresponding end_...() event is received.

This causes the entire element (<foo>) to be sent to the temporary handler ($foo). In the example, each <foo> element will be sent to $foo as a separate document, so if (whitespace shown as underscores)

    <root>
    ____<foo>....</foo>
    ____<foo>....</foo>
    ____<foo>....</foo>
    </root>

is fed in to $d, then $foo will receive 3 separate

    <foo>...</foo>

documents (start_document() and end_document() events are emitted as necessary) and $h will receive a single document without any <foo> elements:

    <root>
    ____
    ____
    ____
    </root>

This can be useful for parsing large XML files in small chunks, often in conjunction with XML::Simple or XML::Filter::XSLT.

But what if you don't want $foo to see three separate documents? What if you're excerpting chunks of a document to create another document? This can be done by telling the dispatcher to emit the main document to $foo and using rules with an action of undef to elide the events that are not wanted. This setup:

    my $foo = XML::Handler::Foo->new();
    my $d = XML::Filter::Dispatcher->new(
        Rules => [
           '/'   => $foo,
           'bar' => undef,
           'foo' => $foo,
        ],
        Handler => $h,
    );

, when fed this document:

    <root>
    __<bar>hork</bar>
    __<bar>
    __<foo>....</foo>
    __<foo>....</foo>
    __<foo>....</foo>
    __<hmph/>
    __</bar>
    __<hey/>
    </root>

results in $foo receiving a single document of input looking like this:

    <root>
    __
    __<foo>....</foo>
    __<foo>....</foo>
    __<foo>....</foo>
    __<hey/>
    </root>

XML::Filter::Dispatcher keeps track of each handler and sends start_document() and end_document() at the appropriate times, so the <foo> elements are "hoisted" out of the <bar> element in this example without any untoward ..._document() events.

TODO: support forwarding to multiple documents at a time. At the present, using multiple handlers for the same event is not supported.

Discrete Event Forwarding

TODO: At the moment, selecting and forwarding individual events is not supported. When it is, any events other than those covered above will be forwarded individually

Tracing

XML::Filter::Dispatcher checks when it is first loaded to see if Devel::TraceSAX is loaded. If so, it will emit tracing messages. Typical use looks like

    perl -d:Devel::TraceSAX script_using_x_f_dispatcher

If you are use()ing Devel::TraceSAX in source code, make sure that it is loaded before XML::Filter::Dispatcher.

TODO: Allow tracing to be enabled/disabled independantly of Devel::TraceSAX.

Namespace Support

XML::Filter::Dispatcher offers namespace support in matching and by providing functions like local-name(). If the documents you are processing don't use namespaces, or you only care about elements and attributes in the default namespace (ie without a "foo:" namespace prefix), then you need not worry about engaging XML::Filter::Dispatcher's namespace support. You do need it if your patterns contain the foo:* construct (that * is literal).

To specify the namespaces, pass in an option like

    Namespaces => {
       ""      => "uri0",   ## Default namespace
       prefix1 => "uri1",
       prefix2 => "uri2",
    },

Then use prefix1: and prefix2: whereever necessary in patterns.

A missing prefix on an element always maps to the default namespace URI, which is "" by default. Attributes are treated likewise, though this is probably a bug.

If your patterns contain prefixes (like the foo: in foo:bar), and you don't provide a Namespaces option, then the element names will silently be matched literally as "foo:bar", whether or not the source document declares namespaces. This may change, as it may cause too much user confusion.

XML::Filter::Dispatcher follows the XPath specification rather literally and does not allow :*, which you might think would match all nodes in the default namespace. To do this, ass a prefixe for the default namespace URI:

    Namespaces => {
       ""        => "uri0",   ## Default namespace
       "default" => "uri0",   ## Default namespace
       prefix1   => "uri1",
       prefix2   => "uri2",
    },

then use "default:*" to match it.

CURRENT LIMITAION: Currently, all rules must exist in the same namespace context. This will be changed when I need to change it (contact me if you need it changed). The current idear is to allow a special function "Namespaces( { .... }, @rules )" that enables a temporary namespace context, although abbreviated forms might be possible.

EventPath Dialect

"EventPath" patterns are that large subset of XPath patterns that can be run in a SAX environment without a DOM. There are a few crucial differences between the environments that EventPath and XPath each operate in.

XPath operates on a tree of "nodes" where each entity in an XML document has only one corresponding node. The tree metaphor used in XPath has a literal representation in memory. For instance, an element <foo> is represented by a single node which contains other nodes.

EventPath operates on a series of events instead of a tree of nodes. For instance elements, which are represented by nodes in DOM trees, are represented by two event method calls, start_element() and end_element(). This means that EventPath patterns may match in a start_...() method or an end_...() method, or even both if you try hard enough.

The only times an EventPath pattern will match in an end_...() method are when the pattern refers to an element's contents or it uses the XXXX function (described below) to do so intentionally.

The tree metaphor is used to arrange and describe the relationships between events. In the DOM trees an XPath engine operates on, a document or an element is represented by a single entity, called a node. In the event streams that EventPath operates on, documents and element

Why EventPath and not XPath?

EventPath is not a standard of any kind, but XPath can't cope with situations where there is no DOM and there are some features that EventPath need (start_element() vs. end_element() processing for example) that are not compatible with XPath.

Some of the features of XPath require that the source document be fully translated in to a DOM tree of nodes before the features can be evaluated. (Nodes are things like elements, attributes, text, comments, processing instructions, namespace mappings etc).

These features are not supported and are not likely to be, you might want to use XML::Filter::XSLT for "full" XPath support (tho it be in an XSLT framework) or wait for XML::TWIG to grow SAX support.

Rather than build a DOM, XML::Filter::Dispatcher only keeps a bare minimum of nodes: the current node and its parent, grandparent, and so on, up to the document ("root") node (basically the /ancestor-or-self:: axis). This is called the "context stack", although you may not need to know that term unless you delve in to the guts.

EventPath Truth

EventPath borrows a lot from XPath including its notion of truth. This is different from Perl's notion of truth; presumably to make document processing easier. Here's a table that may help, the important differences are towards the end:

    Expression      EventPath  XPath    Perl
    ==========      =========  =====    ====
    false()         FALSE      FALSE    n/a (not applicable)
    true()          TRUE       TRUE     n/a
    0               FALSE      FALSE    FALSE
    -0              FALSE**    FALSE    n/a
    NaN             FALSE**    FALSE    n/a (not fully, anyway)
    1               TRUE       TRUE     TRUE
    ""              FALSE      FALSE    FALSE
    "1"             TRUE       TRUE     TRUE

    "0"             TRUE       TRUE     FALSE

 * To be regarded as a bug in this implementation
 ** Only partially implemented/supported in this implementation

Note: it looks like XPath 2.0 is defining a more workable concept for document processing that uses something resembling Perl's empty lists, (), to indicate empty values, so "" and () will be distinct and "0" can be interpreted as false like in Perl. XPath2 is not provided by this module yet and won't be for a long time (patches welcome ;).

EventPath Examples

All of this means that only a portion of XPath is available. Luckily, that portion is also quite useful. Here are examples of working XPath expressions, followed by known unimplemented features.

TODO: There is also an extension function available to differentiate between start_... and end_... events. By default

Examples

 Expression          Event Type      Description (event type)
 ==========          ==========      ========================
 /                   start_document  Selects the document node

 /a                  start_element   Root elt, if it's "<a ...>"

 a                   start_element   All "a" elements

 b//c                start_element   All "c" descendants of "b" elt.s

 @id                 start_element   All "id" attributes

 string( foo )       end_element     matches at the first </foo> or <foo/>
                                     in the current element;
                                     xvalue() returns the
                                     text contained in "<foo>...</foo>"

 string( @name )     start_element   the first "name" attributes;
                                     xvalue() returns the
                                     text of the attribute.

Methods and Functions

There are several APIs provided: general, xstack, and EventPath variable handling.

The general API provides new() and xvalue(), xvalue_type(), and xrun_next_action().

The variables API provides xset_var() and xget_var().

The xstack API provides xadd(), xset(), xoverwrite(), xpush(), xpeek() and xpop().

All of the "xfoo()" APIs may be called as a method or, within rule handlers, called as a function:

    $d = XML::Filter::Dispatcher->new(
        Rules => [
            "/" => sub {
                xpush "foo\n";
                print xpeek;        ## Prints "foo\n"

                my $self = shift;
                print $self->xpeek; ## Also prints "foo\n"
            },
        ],
    );

    print $d->xpeek;                ## Yup, prints "foo\n" as well.

This dual nature allows you to import the APIs as functions and call them using a concise function-call style, or to leave them as methods and use object-oriented style.

Each call may be imported by name:

   use XML::Filter::Dispatcher qw( xpush xpeek );

or by one of three API category tags:

   use XML::Filter::Dispatcher ":general";    ## xvalue()
   use XML::Filter::Dispatcher ":variables";  ## xset_var(), xget_var()
   use XML::Filter::Dispatcher ":xstack";     ## xpush(), xpop(), and xpeek()

or en mass:

   use XML::Filter::Dispatcher ":all";

General API

new
    my $f = XML::Filter::Dispatcher->new(
        Rules => [   ## Order is significant
            "/foo/bar" => sub {
                ## Code to execute
            },
        ],
    );

Must be called as a method, unlike other API calls provided.

xvalue
    "string( foo )" => sub { my $v = xvalue        }, # if imported
    "string( foo )" => sub { my $v = shift->xvalue }, # if not

Returns the result of the last EventPath expression evaluated; this is the result that fired the current rule. The example prints all text node children of <foo> elements, for instance.

For matching expressions, this is equivalent to $_[1] in action subroutines.

xvalue_type

Returns the type of the result returned by xvalue. This is either a SAX event name or "attribute" for path rules ("//a"), or "" (for a string), "HASH" for a hash (note that struct() also returns a hash; these types are Perl data structure types, not EventPath types).

This is the same as xeventtype for all rules that don't evaluate functions like "string()" as their top level expression.

xeventtype

Returns the type of the current SAX event.

xrun_next_action

Runs the next action for the current node. Ordinarily, XML::Filter::Dispatcher runs only one action per node; this allows an action to call down to the next action.

This is especially useful in filters that tweak a document on the way by. This tweaky sort of filter establishes a default "pass-through" rule and then additional override rules to tweak the values being passed through.

Let's suppose you want to convert some mtimes from seconds since the epoch to a human readable format. Here's a set of rules that might do that:

    Rules => [
        'node()' => "Handler",  ## Pass everything through by default.

        'file[@mtime]' => sub { ## intercept and tweak the node.
            my $attr = $_[1]->{Attributes}->{"{}mtime"};

            ## Localize the changes: never assume that it is safe
            ## to alter SAX elements on the way by in a general purpose
            ## filter.  Some smart aleck might send the same events
            ## to another filter with a Tee fitting or even back
            ## through your filter multiple times from a cache.
            local $attr->{Value} = localtime $attr->{Value};

            ## Now that the changes are localised, fall through to
            ## the default rule.
            xrun_next_action;

            ## We could emit other events here as well, but need not
            ## in this example.
         },
    ],

EventPath Variables

EventPath variables may be set in the current context using xset_var(), and accessed using xget_var(). Variables set in a given context are visible to all child contexts. If you want a variable to be set in an enclosed context and later retrieved in an enclosing context, you must set it in the enclosing context first, then alter it in the enclosed context, then retrieve it.

EventPath variables are typed.

EventPath variables set in a context are visible within that context and all enclosed contexts, but not outside of them.

xset_var
    "foo" => sub { xset_var( bar => string => "bingo" ) }, # if imported
    "foo" => sub { shift->xset_var( bar => boolean => 1 ) },

Sets an XPath variables visible in the current context and all child contexts. Will not be visible in parent contexts or sibling contexts.

Legal types are boolean, number, and string. Node sets and nodes are unsupported at this time, and "other" types are not useful unless you work in your own functions that handle them.

Variables are visible as $bar variable references in XPath expressions and using xget_var in Perl code. Setting a variable to a new value temporarily overrides any existing value, somewhat like using Perl's local.

xget_var
    "bar" => sub { print xget_var( "bar" ) }, # if imported
    "bar" => sub { print shift->xget_var( "bar" ) },

Retrieves a single variable from the current context. This may have been set by a parent or by a previous rule firing on this node, but not by children or preceding siblings.

Returns undef if the variable is not set (or if it was set to undef).

xget_var_type
    "bar" => sub { print xget_var_type( "bar" ) }, # if imported
    "bar" => sub { shift->xget_var_type( "bar" ) },

Retrieves the type of a variable from the current context. This may have been set by a parent or by a previous rule firing on this node, but not by children or preceding siblings.

Returns undef if the variable is not set.

Handlers

XML::Filter::Dispatcher allows you to register handlers using set_handler() and get_handler(), and then to refer to them by name in actions. These are part of the "general API".

You may use any string for handler names that you like, including strings with spaces. It is wise to avoid those standard, rarely used handlers recognized by parsers, such as:

    DTDHandler
    ContentHandler
    DocumentHandler
    DeclHandler
    ErrorHandler
    EntityResolver
    LexicalHandler

unless you are using them for the stated purpose. (List taken from XML::SAX::EventMethodMaker).

Handlers may be set in the constructor in two ways: by using a name ending in "Handler" and passing it as a top level option:

    my $f = XML::Filter::Dispatcher->new(
        Handler => $h,
        FooHandler => $foo,
        BarHandler => $bar,
        Rules => [
           ...
        ],
    );

Or, for oddly named handlers, by passing them in the Handlers hash:

    my $f = XML::Filter::Dispatcher->new(
        Handlers => {
            Manny => $foo,
            Moe   => $bar,
            Jack  => $bat,
        },
        Rules => [
           ...
        ],
    );

Once declared in new(), handler names can be used as actions. The "well known" handler name "Handler" need not be predeclared.

For exampled, this forwards all events except the start_element() and end_element() events for the root element's children, thus "hoisting" everything two levels below the root up a level:

    Rules => [
        '/*/*'   => undef,
        'node()' => "Handler",
    ],

By default, no events are forwarded to any handlers: you must send individual events to an individual handlers.

Normally, when a handler is used in this manner, XML::Filter::Dispatcher makes sure to send start_document() and end_document() events to it just before the first event and just after the last event. This prevents sending the document events unless a handler actually receives other events, which is what most people expect (the alternative would be to preemptively always send a start_document() to all handlers when when the dispatcher receives its start_document(): ugh).

To disable this for all handlers, pass the SuppressAutoStartDocument = 1> option.

set_handler
    $self->set_handler( $handler );
    $self->set_handler( $name => $handler );
get_handler
    $self->set_handler( $handler );
    $self->set_handler( $name => $handler );

The xstack

The xstack is a stack mechanism provided by XML::Filter::Dispatcher that is automatically unwound after end_element, end_document, and all other events other than start_element or start_document. This sounds limiting, but it's quite useful for building data structures that mimic the structure of the XML input. I've found this to be common when dealing with data structures in XML and a creating nested hierarchies of objects and/or Perl data structures.

Here's an example of how to build and return a graph:

    use Graph;

    my $d = XML::Filter::Dispatcher->new(
        Rules => [
            ## These two create and, later, return the Graph object.
            'graph'        => sub { xpush( Graph->new ) },
            'end::graph'   => \&xpop,

            ## Every vertex must have a name, so collect in and add it
            ## to the Graph object using its add_vertex( $name ) method.
            'vertex'       => [ 'string( @name )' => sub { xadd     } ],

            ## Edges are a little more complex: we need to collect the
            ## from and to attributes, which we do using a hash, then
            ## pop the hash and use it to add an edge.  You could
            ## also use a single rule, see below.
            'edge'         => [ 'string()'        => sub { xpush {} } ],
            'edge/@*'      => [ 'string()'        => sub { xset     } ],
            'end::edge'    => sub { 
                my $edge = xpop;
                xpeek->add_edge( @$edge{"from","to"} );
            },
        ],
    );

    my $graph = QB->new( "graph", <<END_XML )->playback( $d );
    <graph>
        <vertex name="0" />
        <edge from="1" to="2" />
        <edge from="2" to="1" />
    </graph>
    END_XML

    print $graph, $graph->is_sparse ? " is sparse!\n" : "\n";

should print "0,1-2,2-1 is sparse!\n".

This is good if you can tell what object to add to the stack before seeing content. Some XML parsing is more general than that: if you see no child elements, you want to create one class to contain just character content, otherwise you want to add a container class to contain the child nodes.

An faster alternative to the 3 edge rules relies on the fact that SAX's start_element events carry the attributes, so you can actually do a single rule instead of the three we show above:

            'edge' => sub {
                xpeek->add_edge(
                    $_[1]->{Attributes}->{"{}from"}->{Value},
                    $_[1]->{Attributes}->{"{}to"  }->{Value},
                );
            },
xpush

Push values on to the xstack. These will be removed from the xstack at the end of the current element. The topmost item on the xstack is available through the peek method. Elements xpushed before the first element (usually in the start_document() event) remain on the stack after the document has been parsed and a call like

   my $elt = $dispatcher->xpop;

can be used to retrieve them.

xadd

Tries to add a possibly named item to the element on the top of the stack and push the item on to the stack. It makes a guess about how to add items depending on what the current top of the stack is.

    xadd $name, $new_item;

does this:

    Top of Stack    Action
    ============    ======
    scalar          xpeek                  .= $new_item;
    SCALAR ref      ${xpeek}               .= $new_item;
    ARRAY ref       push @{xpeek()},          $new_item;
    HASH ref        push @{xpeek->{$name}} =  $new_item;
    blessed object  xpeek->$method( $new_item );

The $method in the last item is one of (in order) "add_$name", "push_$name", or "$name".

After the above action, an

    xpush $new_item;

is done.

$name defaults to the LocalName of the current node if it is an attribute or element, so

    xadd $foo;

will DWYM. TODO: search up the current node's ancestry for a LocalName when handling other event types.

If no parameters are provided, xvalue is used.

If the stack is empty, it just xpush()es on the stack.

xset

Like xadd(), but tries to set a named value. Dies if the value is already defined (so duplicate values aren't silently ignored).

    xset $name, $new_item;

does this:

    Top of Stack    Action
    ============    ======
    scalar          xpeek                  = $new_item;
    SCALAR ref      ${xpeek}               = $new_item;
    HASH ref        xpeek->{$name}         = $new_item;
    blessed object  xpeek->$name( $new_item );

Trying to xset any other types results in an exception.

After the above action (except when the top is a scalar or SCALAR ref), an

    xpush $new_item;

is done so that more may be added to the item.

$name defaults to the LocalName of the current node if it is an attribute or element, so

    xset $foo;

will DWYM. TODO: search up the current node's ancestry for a LocalName when handling other event types.

If no parameters are provided, xvalue is used.

xoverwrite

Exactly like xset but does not complain if the value has already been xadd(), xset() or xoverwrite().

xpeek
    Rules => [
        "foo" => sub {
            my $elt = $_[1];
            xpeek->set_name( $elt->{Attributes}->{"{}name"} );
        },
        "/end::*" => sub {
            my $self = shift;
            XXXXXXXXXXXXXXXXXXXX
        }
    ],

Returns the top element on the xstack, which was the last thing pushed in the current context. Throws an exception if the xstack is empty. To check for an empty stack, use eval:

    my $stack_not_empty = eval { xpeek };

To peek down the xstack, use a Perlish index value. The most recently pushed element is index number -1:

    $xpeek( -1 );    ## Same as $self->peek

The first element pushed on the xstack is element 0:

    $xpeek( 0 );

An exception is thrown if the index is off either end of the stack.

xpop
    my $d = XML::Filter::Dispatcher->new(
        Rules => [
            ....rules to build an object hierarchy...
        ],
    );

    my $result = $d->xpop

Removes an element from the xstack and returns it. Usually called in a end_document handler or after the document returns to retrieve a "root" object placed on the stack before the root element was started.

xstack_empty

Handy for detecting a nonempty stack:

    warn xpeek unless xstack_empty;

Because xpeek and xpop throw exceptions on an empty stack, xstack_empty is needed to detect whether it's safe to call them.

xstack_max

Handy for walking the stack:

    for my $i ( reverse 0 .. xstack_max ) {  ## from top to bottom
        use BFD;d xpeek( $i );
    }

Because xpeek and xpop throw exceptions on an empty stack, xstack_max may be used to walk the stack safely.

Notes for XPath Afficianados

This section assumes familiarity with XPath in order to explain some of the particulars and side effects of the incremental XPath engine.

  • Much of XPath's power comes from the concept of a "node set". A node set is a set of nodes returned by many XPath expressions. Event XPath fires a rule once for each node the rule applies to. If there is a location path in the expression, the rule will fire once for each matching event (perhaps twice if both start and end SAX events are trapped, see XXXX below.

    Expressions like 0, false(), 1, and 'a' have no location path and apply to all nodes (including namespace nodes and processing instructions).

  • The XPath parser catches some simple mistakes Perlers might make in typing XPath expressions, such as using && or == instead of and or =.

  • SAX does not define events for attributes; these are passed in to the start_element (but not end_element) methods as part of the element node. XML::Filter::Dispatcher emulates an event for each attribute in order to allow selecting attribute nodes.

  • Axes in path steps (/foo::...)

    Only some axes can be reasonably supported within a SAX framework without building a DOM and/or queueing SAX events for in-document-order delivery.

    On the other hand, lots of SAX-specific Axes are supported.

  • text node aggregation

    SAX does not guarantee that characters events will be aggregated as much as possible, as text() nodes do in XPath. Generally, however, this is not a problem; instead of writing

        "quotation/text()" => sub {
            ## BUG: may be called several times within each quotation elt.
            my $self = shift;
            print "He said '", $self->current_node->{Data}, "'\n'";
        },

    write

        "string( quotation )" => sub {
            my $self = shift;
            print "He said '", xvalue, "'\n'";
        },

    The former is unsafe; consider the XML:

        <quotation>I am <!-- bs -->GREAT!<!-- bs --></quotation>

    Rules like .../text() will fire twice, which is not what is needed here.

    Rules like string( ... ) will fire once, at the end_element event, with all descendant text of quotation as the expression result.

    You can also place an XML::Filter::BufferText instance upstream of XML::Filter::Dispatcher if you really want to use the former syntax (but the GREAT! example will still generate more than one event due to the comment).

  • Axes

    All axes are implemented except for those noted below as "todo" or "not soon".

    Also except where noted, axes have a principal event type of start_element. This node type is used by the * node type test.

    Note: XML::Filter::Dispatcher tries to die() on nonsensical paths like /a/start-document::* or //start-cdata::*, but it may miss some. This is meant to help in debugging user code; the eventual goal is to catch all such nonsense.

    • ancestor:: (XPath, todo, will be limited)

    • ancestor-or-self:: (XPath, todo, will be limited)

    • attribute:: (XPath, attribute)

    • child:: (XPath)

      Selects start_element, end_element, start_prefix_mapping, end_prefix_mapping, characters, comment, and processing_instruction events that are direct "children" of the context element or document.

    • descendant:: (XPath)

    • descendant-or-self:: (XPath)

    • end:: (SAX, end_element)

      Like child::, but selects the end_element event of the element context node.

      This is usually used in preference to end-element:: due to its brevity.

      Because this selects the end element event, most of the path tests that may follow other axes are not valid following this axis. self:: and attribute:: are the only legal axes that may occur to the right of this axis.

    • end-document:: (SAX, end_document)

      Like self::, but selects the end_document event of the document context node.

      Note: Because this selects the end document event, most of the path tests that may follow other axes are not valid following this axis. self:: are the only legal axes that may occur to the right of this axis.

    • end-element:: (SAX, end_element)

      EXPERIMENTAL. This axis is not necessary given end::.

      Like child::, but selects the end_element event of the element context node. This is like end::, but different from end-document::.

      Note: Because this selects the end element event, most of the path tests that may follow other axes are not valid following this axis. attribute:: and self:: are the only legal axes that may occur to the right of this axis.

    • following:: (XPath, not soon)

    • following-sibling:: (XPath, not soon)

      Implementing following axes will take some fucky postponement logic and are likely to wait until I have time. Until then, setting a flag in $self in one handler and checking in another should suffice for most uses.

    • namespace:: (XPath, namespace, todo)

    • parent:: (XPath, todo (will be limited))

      parent/ancestor paths will not allow you to descend the tree, that would require DOM building and SAX event queueing.

    • preceding:: (XPath, not soon)

    • preceding-sibling:: (XPath, not soon)

      Implementing reverse axes will take some fucky postponement logic and are likely to wait until I have time. Until then, setting a flag in $self in one handler and checking in another should suffice for most uses.

    • self:: (XPath)

    • start:: (SAX, start_element )

      This is like child::, but selects the start_element events. This is usually used in preference to start-element:: due to its brevity.

      start:: is rarely used to drive code handlers because rules that match document or element events already only fire code handlers on the start_element event and not the end_element event (however, when a SAX handler is used, such expressions send both start and end events to the downstream handler, so start:: has utility there).

    • start-document:: (SAX, start_document)

      EXPERIMENTAL. This axis is confusing compared to and start-element::, and is not necessary given start::.

      This is like self::, but selects only the start_document events.

    • start-element:: (SAX, start_element)

      EXPERIMENTAL. This axis is not necessary given start::.

      This is like child::, but selects only the start_element events.

  • Implemented XPath Features

    Anything not on this list or listed as unimplemented is a TODO. Ring me up if you need it.

    • String Functions

      • concat( string, string, string* )

      • contains( string, string )

      • normalize-space( string? )

        normalize-space() is equivalent to normalize-space(.).

      • starts-with( string, string )

      • string(), string( object )

        Object may be a number, boolean, string, or the result of a location path:

            string( 10 );
            string( /a/b/c );
            string( @id );

        string() is equivalent to string(.).

      • string-length( string? )

        string-length() not supported; can't stringify the context node without keeping all of the context node's children in mempory. Could enable it for leaf nodes, I suppose, like attrs and #PCDATA containing elts. Drop me a line if you need this (it's not totally trivial or I'd have done it).

      • substring( string, number, number? )

      • substring-after( string, string )

      • substring-before( string, string )

      • translate( string, string, string )

    • Boolean Functions, Operators

      • boolean( object )

        See notes about node sets for the string() function above.

      • false()

      • lang( string ) TODO.

      • not( boolean )

      • true()

    • Number Functions, Operators

      • ceil( number )

      • floor( number )

      • number( object? )

        Converts strings, numbers, booleans, or the result of a location path (number( /a/b/c )).

        Unlike real XPath, this dies if the object cannot be cleanly converted in to a number. This is due to Perl's varying level of support for NaN, and may change in the future.

        number() is equivalent to number(.).

      • round ( number )

      • sum( node-set ) TODO.

    • Node Set Functions

      Many of these cannot be fully implemented in an event oriented environment.

      • last() TODO.

      • position() TODO.

      • count( node-set ) TODO.

    • id( object ) TODO.

    • local-name( node-set? )

    • namespace-uri( node-set? )

    • name( node-set? )

    • All relational operators

      No support for nodesets, though.

    • All logical operators

      Supports limited nodesets, see the string() function description for details.

  • Missing Features

    Some features are entirely or just currently missing due to the lack of nodesets or the time needed to work around their lack. This is an incomplete list; it's growing as I find new things not to implement.

    • count()

      No nodesets => no count() of nodes in a node set.

    • last()

      With SAX, you can't tell when you are at the end of what would be a node set in XPath.

    • position()

      I will implement pieces of this as I can. None are implemented as yet.

  • Todo features

    • id()

    • lang()

    • sum( node-set )

  • Extensions

    • is-start-event(), is-end-event()

      XPath has no concept of time; it's meant to operate on a tree of nodes. SAX has start_element and end_element events and start_document and end_document events.

      By default, XML::Filter::Dispatcher acts on start events and not end events (note that all rules are evaluated on both, but the actions are not run on end_ events by default).

      By including a call to the is-start-event() or is-end-event() functions in a predicate the rule may be forced to fire only on end events or on both start and end events (using a [is-start-event() or is-end-event()] idiom).

TODO

  • Namespace support.

  • Text node aggregation so text() handlers fire once per text node instead of once per characters() event.

  • Nice messages on legitimate but unsupported axes.

  • /../ (parent node)

  • add_rule(), remove_rule(), set_rules() methods.

OPTIMIZING

Pass Assume_xvalue => 0 flag to tell X::F::D not to support xvalue and xvalue_type, which lets it skip some instructions and run faster.

Pass SortAttributes => 0 flag to prevent calling sort() for each element's attributes (note that Perl changes hashing algorithms occasionally, so setting this to 0 may expose ordering dependancies in your code).

DEBUGGING

NOTE: this section describes things that may change from version to version as I need different views in to the internals.

Set the option Debug => 1 to see the Perl code for the compiled ruleset. If you have GraphViz.pm and ee installed and working, set Debug => 2 to see a graph diagram of the intermediate tree generated by the compiler.

Set the env. var XFDSHOWBUFFERHIGHWATER=1 to see what events were postponed the most (in terms of how many events had to pile up behind them). This can be of some help if you experience lots of buffering or high latency through the filter. Latency meaning the lag between when an event arrives at this filter and when it is dispatched to its actions. This will only report events that were actually postponed. If you have a 0 latency filter, the report will list no events.

Set the env. var XFDOPTIMIZE=0 to prevent all sorts of optimizations.

LIMITATIONS

  • NaN is not handled properly due to mediocre support in perl, especially across some platforms that it apparently isn't easily supported on.

  • -0 (negative zero) is not provided or handled properly

  • +/- Infinity is not handled properly due to mediocre support in perl, especially across some platforms that it apparently isn't easily supported on.

This is more of a frustration than a limitation, but this class requires that you pass in a type when setting variables (in the Vars ctor parameter or when calling xset_var). This is so that the engine can tell what type a variable is, since string(), number() and boolean() all treat the Perlian 0 differently depending on its type. In Perl the digit 0 means false, 0 or '0', depending on context, but it's a consistent semantic. When passing a 0 from Perl lands to XPath-land, we need to give it a type so that string() can, for instance, decide whether to convert it to '0' or 'false'.

THANKS

...to Kip Hampton, Robin Berjon and Matt Sergeant for sanity checks and to James Clark (of Expat fame) for posting a Yacc XPath grammar where I could snarf it years later and add lots of Perl code to it.

AUTHOR

    Barrie Slaymaker <barries@slaysys.com>

COPYRIGHT

    Copyright 2002, Barrie Slaymaker, All Rights Reserved.

You may use this module under the terms of the Artistic or GNU Pulic licenses your choice. Also, a portion of XML::Filter::Dispatcher::Parser is covered by:

        The Parse::Yapp module and its related modules and shell scripts are
        copyright (c) 1998-1999 Francois Desarmenien, France. All rights
        reserved.

        You may use and distribute them under the terms of either the GNU
        General Public License or the Artistic License, as specified in the
        Perl README file.

Note: Parse::Yapp is only needed if you want to modify lib/XML/Filter/Dispatcher/Grammar.pm